jupyter nbconvert --to slides cleanwater.ipynb --post serve --SlidesExporter.reveal_scroll=True cd Documents/GitHub/Annual_Water_Depletion/dataproject_water/

OpenDataProject - CODE1161 ⭐✨

Clean Water & Sanitation 🌏💧

Htay Hlaing Win - z5443603

---

Looking at our beautiful blue planet from space, water appears abundant, covering about 71% of the Earth's surface. However, there's a hidden truth that often goes unnoticed. While water might seem endless, the availability of freshwater, the life-sustaining resource we rely on for our survival, is incredibly limited.

from IPython.display import Image
Image(filename="water_pic.png")

Freshwater is a limited resource

---

Out of all the water on our planet, only a mere 2.5 percent is freshwater. The remaining 97.5 percent is Non-freshwater , making it unsuitable for direct human consumption or crop irrigation.

Did you know that about 69 percent of all freshwater is actually trapped in glaciers? That means it's not readily available for use. From the chart, Groundwater makes up the largest portion of the remaining freshwater. Surprisingly, it serves as the primary source of water supplies in many countries, even though it is hidden beneath the Earth's surface and can only be accessed through drilling.The solid state of ice caps, glaciers, and permanent snow makes them difficult to access for obtaining water. Thus, easily accessible surface freshwater accounts for less than 1% of the total.

Water Availability 🌊🚰

---

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter, MultipleLocator

#format tick labels as thousands (K)
def thousands_formatter(x, pos):
    return f'{x / 1e3:.0f}K'

#format tick labels as millions
def millions_formatter(x, pos):
    return f'{x / 1e6:.0f}M'

data = pd.read_csv("Per_capita2.csv")
countries_to_plot = ['Australia']

# Create a figure and axis
for country in countries_to_plot:
    plt.figure(figsize=(14, 8))   
    # Filter data for the current country starting from 1965
    country_data = data[(data['country'] == country) & (data['year'] >= 1965)]    
    # Bar Chart
    bar = plt.bar(country_data['year'], country_data['per_capita'], color='grey', label='Per Capita')
    plt.xlabel('Year')
    plt.ylabel('Water Sources Per Capita (in billion m³)')
    plt.title(f'Water Availability Over Years - {country}')
    plt.grid(True)    
    # Customize y-axis tick labels
    plt.gca().yaxis.set_major_formatter(FuncFormatter(thousands_formatter)) 

    # Line Chart 
    ax2 = plt.gca().twinx()
    line = ax2.plot(country_data['year'], country_data['population'], marker='o', color='red', linestyle='-', linewidth=2, markersize=6, label='Population')
    ax2.set_ylabel('Population (Millions)')

    # Set y-axis tick label formatter for the line chart
    ax2.yaxis.set_major_formatter(FuncFormatter(millions_formatter))  
    
    # Set x-axis tick interval to every 5 years
    ax2.xaxis.set_major_locator(MultipleLocator(base=5))

    # Get handles and labels for the line chart legend only
    line_handles, line_labels = ax2.get_legend_handles_labels()

    # Show legend for the line chart
    plt.legend(handles=line_handles, labels=line_labels, loc='upper left', bbox_to_anchor=(0.75, 1.0))

    plt.tight_layout()
    plt.show()

Next I will showcase the insightful data about water availabiliy in Aus from 1965. In this graph, I have combined bar chart and line chart together. customized the y-axis tick labels to be more readable, using thousands formatting on the left and million formatting on the right. Each bar represents a year, and the height indicates water sources available to each person in a country in billion cubic meter. By looking at the chart together, we can see the trends in water availability and population growth in aus over time. The bar chart allows us to visualize changes in water sources per capita, while the line chart offers insights into shifts in population dynamics.

import pandas as pd
data = pd.read_csv("Per_capita2.csv")
per_capita_difference = data.groupby('country').apply(lambda x: x['per_capita'].iloc[-1] - x['per_capita'].iloc[0])
sorted_countries = per_capita_difference.sort_values(ascending=False)
print("Countries with the biggest decrease in per capita water resources:")
print(sorted_countries.head())

Countries with the biggest decrease in per capita water resources:
country
Congo, Rep.         424206.10940
Iceland             399686.47630
Bhutan              283583.85120
Solomon Islands     271623.84635
Papua New Guinea    239530.79763
dtype: float64

import pandas as pd
data = pd.read_csv("Per_capita2.csv")
per_capita_difference = data.groupby('country').apply(lambda x: x['per_capita'].iloc[-1] - x['per_capita'].iloc[0])
sorted_countries = per_capita_difference.sort_values()
countries_with_biggest_increase = sorted_countries.head(3).index
print("Countries with the biggest increase in per capita water resources:")
print(countries_with_biggest_increase)

Countries with the biggest increase in per capita water resources:
Index(['Latvia', 'Georgia', 'Croatia'], dtype='object', name='country')

import pandas as pd
import matplotlib.pyplot as plt

# Load the data from the CSV file
data = pd.read_csv("Per_capita2.csv")

# Calculate the per capita difference for each country
per_capita_difference = (data.groupby('country')['per_capita'].last()
                         - data.groupby('country')['per_capita'].first())

# Sort the countries based on the per capita difference
sorted_countries = per_capita_difference.sort_values()

# Plot the data
plt.figure(figsize=(10, 6))
sorted_countries.plot(kind='bar', color='red')
plt.xlabel('Country')
plt.ylabel('Per Capita Difference')
plt.title('Countries with the Biggest Decrease in Per Capita Water Resources')
plt.xticks(rotation=45, ha='right')
plt.tight_layout()
plt.show()

What about the world. Most of the countries are not doing great in maintaining available water sources over time.

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter
data = pd.read_csv("Per_capita2.csv")

per_capita_difference = data.groupby('country').apply(lambda x: x['per_capita'].iloc[-1] - x['per_capita'].iloc[0])
label_prefixes = ['First', 'Second', 'Third']

top_decrease_countries = per_capita_difference.sort_values().head(3).index
smallest_decrease_countries = per_capita_difference.sort_values(ascending=False).head(3).index

# Create a function to format tick labels as thousands (K)
def thousands_formatter(x, pos):
    return f'{x / 1e3:.0f}K'
plt.figure(figsize=(10, 6))

top_increase_colors = ['green', 'olive', 'black']
top_decrease_colors = ['red', 'orange', 'yellow']

for i, country in enumerate(top_decrease_countries):
    country_data = data[data['country'] == country]
    plt.scatter(country_data['year'], country_data['per_capita'], color=top_increase_colors[i], label=f'{label_prefixes[i]} Biggest Increase: {country}')

for i, country in enumerate(smallest_decrease_countries):
    country_data = data[data['country'] == country]
    plt.scatter(country_data['year'], country_data['per_capita'], color=top_decrease_colors[i], 
                label=f'{label_prefixes[i]} Biggest Decrease: {country}')

plt.xlabel('Year')
plt.ylabel('Water Sources Per Capita (in billion m³)')
plt.title('Water Availability Over Years - Per Capita Differences')

# Apply the custom thousands formatter to y-axis
plt.gca().yaxis.set_major_formatter(FuncFormatter(thousands_formatter))
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.show()

Water Withdrawls 🌍💧

---

import matplotlib.pyplot as plt
import pandas as pd
import os
import seaborn as sns

if os.path.isfile("water_withdrawal_by_sector.csv"):
    filepath = "water_withdrawal_by_sector.csv"
    print("Loading from file")
else:
    filepath = "https://datatopics.worldbank.org/sdgatlas/data/goal06/goal06.water_withdrawal_by_sector.csv"
    print("Loading from the internet")

water_usage_data = pd.read_csv(filepath)

income_categories = ["HIC", "LIC", "LMC", "UMC"]
sectors = ['agriculture', 'domestic', 'industry']

# List of region names to be removed from x-axis
regions_to_remove = ['East Asia & Pacific', 'Europe & Central Asia', 'Latin America & Caribbean',
                     'Middle East & North Africa', 'North America', 'South Asia', 
                     'Sub-Saharan Africa', 'World']

# Filter out specific regions from the data
filtered_data = water_usage_data[~water_usage_data['region_name'].isin(regions_to_remove)]

# Melt the data to create a suitable format for Seaborn
melted_data = pd.melt(filtered_data, id_vars=['region_name', 'region'], value_vars=sectors, var_name='sector')

# Set the style
sns.set(style="whitegrid")

# Create a grouped bar chart
plt.figure(figsize=(10, 6))
sns.barplot(data=melted_data, x='region_name', y='value', hue='sector', palette='Set2')

plt.title('Water Withdrawal by Sectors for Different Income Categories')
plt.ylabel('Percentage')
plt.xlabel('Countries')
plt.xticks(rotation=45)

plt.legend(title='Sector')
plt.tight_layout()
plt.show()

Loading from file

This graph visually captures how countries of varying income categories allocate their water resources across sectors like agriculture, domestic use, and industry. It turns out that the agriculture sector is the real heavyweight when it comes to water usage, no matter the income level. Now, in countries with higher incomes, people tend to use a lot of water for their homes and industries. On the flip side, low-income countries don't use as much water for domestic and industrial purposes. Instead, they use most water for agriculture. And the interesting thing is low income and lower middle income countries have similar usage of water. However, this similarity extends to high-income and higher middle-income nations as well, suggesting a surprising commonality in water resource management strategies across different income level

import matplotlib.pyplot as plt
import pandas as pd
import os
import seaborn as sns

if os.path.isfile("water_withdrawal_by_sector.csv"):
    filepath = "water_withdrawal_by_sector.csv"
    print("Loading from file")
else:
    filepath = "https://datatopics.worldbank.org/sdgatlas/data/goal06/goal06.water_withdrawal_by_sector.csv"
    print("Loading from the internet")

percent_data = pd.read_csv(filepath)

# Exclude the rows
filtered_data = percent_data[~percent_data['region'].isin(['HIC', 'LIC','LMC', 'UMC'])]

# Set the style
sns.set(style="whitegrid")

# Create a stacked area chart for different sectors over time
sectors = ['agriculture', 'domestic', 'industry']
plt.figure(figsize=(10, 6))
sns.lineplot(data=filtered_data, x='region_name', y=sectors[0], label=sectors[0])
sns.lineplot(data=filtered_data, x='region_name', y=sectors[1], label=sectors[1])
sns.lineplot(data=filtered_data, x='region_name', y=sectors[2], label=sectors[2])

plt.fill_between(filtered_data['region_name'], filtered_data[sectors[0]], color='blue', alpha=0.3)
plt.fill_between(filtered_data['region_name'], filtered_data[sectors[1]], color='green', alpha=0.3)
plt.fill_between(filtered_data['region_name'], filtered_data[sectors[2]], color='orange', alpha=0.3)

plt.title('Water Withdrawal by Sectors')
plt.ylabel('Percentage')
plt.xlabel('Region')
plt.xticks(rotation=45)
plt.legend(title='Sector')

plt.tight_layout()
plt.show()

Loading from file

Water Stress ⚠️🌎

---

The water stress is calculated by dividing the total freshwater withdrawals by the available renewable freshwater.As a country's water usage approaches the 100 percent,he pressure on its water resources intensifies. In some cases, the value of 100 percent can even be exceeded. This happens when a country consumes water resources that are non-renewable. While low water stress indicates sustainable usage, it is important to note that this is not necessarily a positive outcome:. It can sometimes highlight a nation's struggle to effectively govern its water resources for the benefit of its population. In some low-income countries, water stress levels are very low, but access levels are also low

import geopandas as gpd
import plotly.express as px
import pandas as pd

water_stress_data = pd.read_csv('country_level_water_stress.csv')
shapefile_path = 'ne_110m_admin_0_countries/ne_110m_admin_0_countries.shp'
world = gpd.read_file(shapefile_path)
# Choose the correct ISO country code column name
correct_iso_column_name = 'ADM0_A3'

# Merge shapefile with water stress data based on the correct ISO column name
merged_data = world.merge(water_stress_data, left_on=correct_iso_column_name, right_on='iso', how='left')

# Create an interactive map with hover effect using plotly express
fig = px.choropleth(merged_data,
                    geojson=merged_data.geometry,
                    locations=merged_data.index,
                    color='water_stress',
                    color_continuous_scale='OrRd',
                    hover_name='ADMIN',  # Include country names in hover
                    hover_data={'water_stress': ':.2f'},
                    title='Water Stress in Countries',
                    labels={'water_stress': 'Water Stress (%)'},
                    color_continuous_midpoint=100,  # Set midpoint for color scale
                    range_color=(0, 100))  # Set range of colors
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(height=600)
fig.show()

If we focus on the Middle East and North Africa region, you'll notice that countries there are dealing with critical or high water stress. Why? Well, it's because water resources are pretty scarce in that area. The population of these countries has nearly quadrupled over the past 50 years and is projected to keep increasing. The growing demand for water puts pressure on the Nile River Basin, where water flows are already fully used to meet various domestic, agricultural and industrial demands.

import pandas as pd
import matplotlib.pyplot as plt

data = pd.read_csv("Per_capita2.csv")
countries_to_plot = ['Libya', 'Egypt, Arab Rep.', 'Algeria', 'Saudi Arabia']  # Change to the desired countries

plt.figure(figsize=(14, 8))

# Create a histogram for each country's population
for country in countries_to_plot:
    # Filter data for the current country for years 1965 and 2020
    country_data = data[(data['country'] == country) & (data['year'].isin([1965, 2020]))]
    
    # Create a list of labels with years and countries
    labels = [f'{year} - {country}' for year in country_data['year']]
    
    # Vertical bar chart (swapping y and x axes)
    plt.bar(labels, country_data['population'] / 1e6, alpha=0.5, label=f'{country} Population')

# Customize the plot
plt.xlabel('Years - Countries')
plt.ylabel('Population (Millions)')
plt.title('Population Distribution by Country and Year')
plt.grid(True)
plt.legend()

plt.tight_layout()
plt.show()

Looking Forward ↗️🔱

---

Water is a precious resource. Amid pressures from human activities, and uncertainties linked to climate change, robust water management will be critical to the sustainable use of freshwater resources. This is especially important when considering the interdependencies that exist between water, food production, energy, ecosystems, and economic activities. Thus, lets the first step towards a water-efficient future by integrating easy water-saving habits into our daily lives. Thank you. 💧🌿